Table 1.

Ethical standards to be followed when evaluating and developing AI in nuclear medicine (adapted from Currie et al.)

1	Beneficence, i.e. common good
2	Non-maleficence, i.e. do no harm
3	Fairness and justice, i.e. equal opportunity and access
4	Safety, i.e. for the patient
5	Reliability, i.e. accuracy and reproducibility in clinical practice
6	Security, i.e. for the data
7	Privacy and confidentiality of data
8	Mitigation of bias, i.e. fair and evidence-based clinical validation
9	Transparency and visibility, i.e. to the patients and community
10	Explainability and comprehensibility
11	Human values: human-in-the-loop process incorporated
12	Autonomy, judgement, and decision-making: human-in-the-loop process
13	Collegiality, i.e. multidisciplinary involvement and commitment
14	Accountability, i.e. among stakeholders
15	Governance, i.e. framework to ensure compliance
16	Inclusiveness, i.e. empowerment of all stakeholders